chore(rivetkit): remove wasm sqlite runtime

[NathanFlurry]

Description

Please include a summary of the changes and the related issue. Please also include relevant motivation and context.

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[NathanFlurry]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• chore: lockfile lefthook #4647
• fix(rivetkit): surface native sqlite kv errors to callers #4645
• fix(rivetkit): prevent sleep races during disconnect and db work #4644
• fix(rivetkit): isolate engine envoys and propagate startup failures #4643
• test(rivetkit): remove wasm sqlite-only fixtures and coverage #4642
• chore(sqlite-vfs): remove wasm sqlite packages and workspace wiring #4641
• chore(rivetkit): remove wasm sqlite runtime #4614 👈 (View in Graphite)
• fix(sqlite-native): restore native startup kv preload #4639
• fix(sqlite-native): delete metadata before chunk range #4638
• fix(sqlite-native): keep truncate cache coherent #4637
• fix(sqlite-vfs): use delete range for truncate cleanup #4636
• fix(sqlite-native): restore kv error hook #4635
• perf(sqlite-native): avoid cloning cached read chunks #4634
• perf(sqlite-native): gate kv operation labels #4633
• perf(sqlite-native): reuse read cache for partial writes #4632
• perf(sqlite-native): remove delete file existence probe #4631
• fix(sqlite-native): gate native read cache behind env var #4630
• fix(pegboard): isolate runner config dc lookup failures #4625 : 1 other dependent PR (#4626 )
• docs(pegboard): note runner config upsert split write risk #4624
• fix(pegboard): unpack global runner config keys correctly #4623
• fix(pegboard-gateway): enforce tunnel message state #4622
• fix(pegboard-runner): clear terminal tunnel routes #4621
• fix(pegboard): skip protocol version keys in runner pool backfill #4620
• fix(pegboard): create missing normal runner configs #4619
• fix(epoxy): reuse cached optimistic value #4618
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[claude]

PR Review: chore(rivetkit): remove wasm sqlite runtime

This PR removes the WASM/KV-backed SQLite path from RivetKit, making native SQLite via @rivetkit/rivetkit-native the only supported backend. The simplification is substantial: ~1,700 lines deleted, eliminating the WASM pool manager, KV-backed VFS, native-sqlite.ts KV channel module, native-adapter.ts, open-database.ts, and createActorKvStore. The new SqliteDatabase interface in config.ts and wrapJsNativeDatabase in native-database.ts are cleaner than the previous fallback chain.

---

Bug: ArrayBuffer and typed array support dropped

native-database.ts:toNativeBinding only handles Uint8Array, but the previous wrapper.js handled all ArrayBuffer views:

// old wrapper.js — handled both:
if (value instanceof ArrayBuffer) { ... Buffer.from(value) }
if (ArrayBuffer.isView(value)) { ... Buffer.from(value.buffer, value.byteOffset, value.byteLength) }

The new code only handles Uint8Array, so Int8Array, Float32Array, bare ArrayBuffer, etc. will throw. This is a silent regression if any caller passes those types.

---

Bug: bigint precision loss in native-database.ts

toNativeBinding uses Number(arg) for bigints, silently losing precision for values outside Number.MAX_SAFE_INTEGER (2^53 - 1). SQLite stores integers up to i64, so this is reachable. Consider throwing when the value exceeds safe integer range.

---

Edge case: named params with no placeholders

In toNativeBindings (native-database.ts), when a Record<string, unknown> is passed but the SQL has no named placeholders (orderedNames.length === 0), it falls through to Object.values(params) used positionally. The old native-adapter.ts threw an explicit error here. Using Object.values positionally is fragile — throwing would be safer and matches the old behavior.

---

Orphaned KV data

SQLITE_PREFIX (key byte 0x08) is removed from keys.ts with no cleanup or migration. Actors that previously used the KV-backed WASM SQLite path have data silently orphaned under that prefix. Confirm whether this is an intentional hard cutover and document it in the PR description — existing actors would start with an empty database.

---

Removed error poisoning — verify native layer handles it

The old path poisoned the KV store before acquiring the mutex on close() so in-flight queries would fail fast with a descriptive message. This mechanism is now gone. Confirm the native backend surfaces a clear error (not a hang) when the database is closed mid-query.

---

Static import of openDatabaseFromEnvoy

The engine driver now statically imports openDatabaseFromEnvoy at module level with no fallback. The old code used a runtime require() with a try/catch returning undefined if unavailable. Deployments without @rivetkit/rivetkit-native will now fail at import time rather than gracefully. Confirm this package is always present in all deployment targets.

---

Minor: duplicate interface

NativeExecBridgeResult in isolate-runtime.ts duplicates SqliteQueryResult from db/config.ts. Could reuse the existing type.

---

Overall

The architectural direction is correct — eliminating the WASM/KV complexity in favor of native-only is a clear win for maintainability. The dynamic actor DB bridge (dbExec, dbQuery, dbRun, dbClose refs) is well-structured and consistent with the existing KV bridge pattern.

Items to address before merging:

1. ArrayBuffer regression — restore ArrayBuffer and ArrayBuffer.isView handling
2. bigint precision — silent data corruption for integers > 2^53
3. Named params edge case — throw instead of falling through to Object.values positionally
4. Orphaned KV data — document the intentional cutover
5. Static import — confirm @rivetkit/rivetkit-native is always present in all deployment targets

[github-actions]

Preview packages published to npm

Install with:

npm install rivetkit@pr-4614

All packages published as 0.0.0-pr.4614.822b57f with tag pr-4614.

Engine binary is shipped via @rivetkit/engine-cli (platforms: linux-x64-musl, linux-arm64-musl, darwin-x64, darwin-arm64). rivetkit resolves it automatically at runtime.

Docker images:

docker pull rivetdev/engine:slim-822b57f
docker pull rivetdev/engine:full-822b57f

Individual packages

npm install rivetkit@pr-4614
npm install @rivetkit/react@pr-4614
npm install @rivetkit/rivetkit-native@pr-4614
npm install @rivetkit/workflow-engine@pr-4614